International Journal of Pharmaceutics 193 (2000) 137–146
`
`www.elsevier.com:locate:ijpharm
`
`Physicochemical properties and bioavailability of
`carbamazepine polymorphs and dihydrate
`
`Yumiko Kobayashi a,*, Shusei Ito a, Shigeru Itai a, Keiji Yamamoto b
`a Research Center, Taisho Pharmaceutical Company Ltd, 1-403 Yoshinocho, Omiya, Saitama 330-8530, Japan
`b Faculty of Pharmaceutical Sciences, Chiba Uni6ersity, 1-33 Yayoicho, Inageku, Chiba 263-8522, Japan
`
`Received 19 May 1999; received in revised form 6 August 1999; accepted 30 August 1999
`
`Abstract
`
`The dissolution behaviors of carbamazepine (CZP) polymorphs and pseudopolymorphs (form I, form III and
`dihydrate) and the bioavailabilities (BA) of each form in dogs after oral administration were investigated.
`Bioavailability tests were carried out at a dose of either 40 mg:body or 200 mg:body. The results of dissolution tests
`in JP13 first fluid (pH 1.2) at 37°C indicated that the initial dissolution rate was in the order of form III\form
`I\dihydrate, while form III was transformed to dihydrate more rapidly than form I, resulting in decrease of the
`dissolution rate. The solubilities of both anhydrates (form I and form III), calculated from the initial dissolution rate
`of each anhydrate, were 1.5–1.6 times that of the dihydrate. At the dose of 40 mg:body, there were no significant
`differences in the area under the curve (AUC) between forms; their AUCs were nearly equal to that of CZP solution
`using polyethyleneglycol 400. These findings suggested that most crystalline powder of each form administered at the
`low dose was rapidly dissolved in gastrointestinal (GI) fluid. On the other hand, for the dose of 200 mg:body,
`significant differences in plasma concentration–time curves of CZP among polymorphic forms and dihydrate were
`observed. The order of AUC values was form I\form III\dihydrate. The inconsistency between the order of initial
`dissolution rates and that of AUC values at the high dose may have been due to rapid transformation from form III
`to dihydrate in GI fluids. © 2000 Published by Elsevier Science B.V. All rights reserved.
`
`Keywords: Carbamazepine; Polymorph; Dihydrate; Dissolution rate; Transformation; Bioavailability
`
`1. Introduction
`
`For drugs which have several polymorphs or
`pseudopolymorphs, differences in bioavailability
`(BA) between forms have been reported (York,
`1983; Rajendra and David, 1995). Carbamazepine
`
`* Corresponding author. Tel.: (cid:27)81-48-6631111; fax: (cid:27)81-
`48-6527254.
`
`(CZP), which has at least four polymorphic forms
`and a dihydrate (Kaneniwa et al., 1984; Krahn
`and Mielck, 1987), is a widely prescribed anticon-
`vulsant antiepileptic drug. Meyer et al. (1992)
`compared the BAs of three lots of a generic 200
`mg CZP tablet to that of one lot of the innovator
`product in healthy volunteers and found signifi-
`cant differences in the rate and extent of absorp-
`tion between the generic products and the
`
`0378-5173:00:$ - see front matter © 2000 Published by Elsevier Science B.V. All rights reserved.
`PII: S 0 3 7 8 - 5 1 7 3 ( 9 9 ) 0 0 3 1 5 - 4
`
`Merck Exhibit 2162, Page 1
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`138
`
`Y. Kobayashi et al. :International Journal of Pharmaceutics 193 (2000) 137–146
`
`innovator product, as well as among the generic
`lots. The mean maximum CZP plasma concentra-
`tions for two of the generic lots were only 61–
`74% of that of the innovator product, while that
`for the third lot was 142% of that of the innovator
`product. However,
`it has not been clear what
`causes the differences in bioavailability among
`CZP tablets. Some factors contribute to these
`differences, i.e. differences in crystalline form and:
`or particle size of CZP raw material.
`Behme and Brooke (1991) reported that CZP
`form I (USP grade material) and form III, which
`melt at 176 and 189°C, respectively, were enan-
`tiotropic and had a transition temperature at
`71°C.
`Kaneniwa et al. (1984) demonstrated that when
`form I and form III were stored under water
`vapor condition at 37°C for 2 weeks, both crystal
`forms were transformed to dihydrate. Kaneniwa
`et al. (1987) carried out a dissolution study of the
`two anhydrates (form I and form III) and dihy-
`drate in water using the rotating disk method.
`The dissolution study revealed that initial dissolu-
`tion rates of both anhydrates were higher than
`that of the dihydrate and the anhydrates trans-
`formed to the dihydrate rapidly. However, in their
`study, the initial dissolution rate of form I was
`higher than that of form III, even though their
`experiment was performed in the temperature
`range in which form I was more stable than form
`III according to the theory of thermodynamic
`stability.
`The bioavailabilities of one anhydrate (crystal
`form was not shown) and dihydrate were investi-
`gated by Kahela et al. (1983). In a comparison of
`the anhydrate and dihydrate at a dose of 200
`mg:body in humans after oral administration,
`there was no marked difference between the
`plasma concentration–time curves of
`the two
`crystal forms. Whereas in the bioavailability test
`of generic 200 mg CZP tablets studied by Meyer
`et al. (1992) described above, the notable differ-
`ences between the area under the curve (AUC)
`values of tablets were observed. Various studies of
`physicochemical properties of form I, form III
`and the dihydrate have been reported; however,
`there are some discrepancies among findings of
`these studies, and the relationship between physic-
`
`ochemical properties of CZP polymorphic forms
`and dihydrate and the BA of CZP is not com-
`pletely understood.
`In the present study, the dissolution properties
`of form I, form III and dihydrate and the behav-
`iors of transformation from form I or form III to
`the dibydrate during dissolution tests were investi-
`gated and bioavailability tests in dogs were per-
`formed in order to determine the effects of
`physicochemical properties of form I, form III
`and dihydrate on the plasma level of CZP. The
`bioavailability tests in this study were carried out
`at a dose of either 40 mg:body or 200 mg:body,
`since the absorption of drugs with poor solubility
`such as CZP were
`affected by
`the dose
`administered.
`
`2. Materials
`
`Carbamazepine was obtained from Wako Pure
`Chemical Industries (Japan; sample A). Other
`crystalline forms were obtained according to the
`method described by McMahon et al. (1996).
`Sample B was prepared by heating sample A at
`170°C for 2 h. Sample C was prepared by sus-
`pending sample A in distilled water for 24 h at
`room temperature, then dried on filter at room
`temperature for 30 min. Sample B was ground
`using an agate centrifugal ball mill (Model Pul-
`verisette 5, Fritsch) for 5 min (sample B%), since
`the particle size of sample B was significantly
`greater than those of samples A and C (Fig. 1).
`The mean particle size (d) was determined using
`the microscopic technique (Microscope; E8-21-1,
`Nikon; Real-time image analyzer; Luzex- F,
`Nireco) as Heywood diameter, and the specific
`surface area (S) was determined by the air perme-
`ability method (Powder specific surface area me-
`ter; Model SS-100, Shimadzu) (Table 1).
`
`3. Methods
`
`3.1. Identification of crystalline forms of samples
`
`3.1.1. Powder X-ray diffractometry
`The powder X-ray diffraction patterns were
`
`Merck Exhibit 2162, Page 2
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`Y. Kobayashi et al. :International Journal of Pharmaceutics 193 (2000) 137–146
`
`139
`
`determined with an X-ray diffractometer (Model
`RAD3-C, Rigaku). The conditions of measure-
`ment were as follows: Target; Cu, filter; Ni,
`voltage; 40 kV, current; 30 mA, scanning speed;
`4°:min, scanning angle; 3(cid:2)40°.
`
`3.1.2. Differential scanning calorimetry (DSC)
`DSC curves were obtained with a differential
`scanning calorimeter
`(Model DSC-7, Perkin-
`Elmer). Differential scanning calorimetry was per-
`formed under the following conditions: Sample
`weight; about 2 mg, sample cell; an aluminium
`open cell with a cell cover, nitrogen flow rate; 20
`ml:min, heating rate; 10°C:min.
`
`Fig. 1. Microscopic photographs of CZP samples. a: Sample
`A, b: Sample B, c: Sample C, d: Sample B%.
`
`Table 1
`Mean particle diameter and specific surface area of samples
`
`Mean particle di-
`ameter (d) (mm)
`
`Specific surface area
`(S) (cm2:g)
`
`Sample A
`Sample B
`Sample B%
`Sample C
`
`13.9
`108.0
`19.5
`13.4
`
`1.27(cid:29)103
`2.43(cid:29)102
`1.10(cid:29)103
`1.24(cid:29)103
`
`3.1.3. Thermogra6imetric analysis (TG)
`TG curve was obtained with a thermogravimet-
`ric analyzer (Model TGA-7, Perkin-Elmer). Ther-
`mogravimetry was performed under the following
`conditions: Sample weight; about 8 mg, sample
`cell; a platinum open cell, nitrogen flow rate; 70
`ml:min, heating rate; 10°C:min.
`
`3.2. Physicochemical properties
`
`3.2.1. Dissolution studies by the static disk method
`The intrinsic dissolution rates of samples A, B%
`and C were determined by the static disk method
`described in the previous report (Ito et al., 1997).
`It was confirmed by powder X-ray diffraction
`analysis that no polymorphic transition took
`place during disk preparation for any sample.
`Four hundred milliliters of JP13 first fluid (pH
`1.2) at 37°C was used as the dissolution medium,
`which was stirred at 150 rpm with a paddle. At
`definite time intervals, the solution was passed
`through G-3 glass filter and delivered to the cell
`using pump attached to the apparatus. The con-
`centration of CZP in the solution was determined
`by measurement of the absorbance at 285nm (Ul-
`traviolet spectrophotometer; Model W-1600, Shi-
`madzu). The sampling solution was returned to
`the original solution by the circulation system.
`
`3.2.2. Dissolution studies by the dispersion method
`The dissolution behaviors of CZP samples in
`the JP13 first fluid at 37°C were investigated by
`the dispersion method. Approximately 50 mg of
`each sample was added to 30 ml of JP13 first fluid
`maintained at 37°C and the suspension was
`stirred at 650 rpm with a stirrer (Model RCN-7R,
`Eyela) and sampled periodically. After filtration
`(pore size 0.45 mm, Gelman Sciences) of sampling
`solution, the concentration of CZP was deter-
`mined by high performance liquid chromatogra-
`phy
`(HPLC). High
`performance
`liquid
`chromatography analysis was performed using a
`Shimadzu HPLC chromatograph composed of an
`LC10-AT and SPD-lOAV. The conditions of
`HPLC method were as follows: Mobile phase;
`acetonitrile:0.1 M ammonium acetate(cid:30)270:
`
`Merck Exhibit 2162, Page 3
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`140
`
`Y. Kobayashi et al. :International Journal of Pharmaceutics 193 (2000) 137–146
`
`3.3.2. Preparation of capsules and solution
`Capsules: Each sample was mixed with lactose
`at a weight ratio of 1:2 (CZP: lactose) and the
`mixture was placed in the hard gelatin capsule.
`c
`c
`Forty milligram and 100 mg of CZP, as anhy-
`drate, were contained in
`3 and
`1 capsules,
`respectively. Solution: 40 mg or 200 mg of sample
`A were dissolved in 30 ml of PEG 400.
`
`3.3.3. Determination of CZPconcentration in
`plasma
`standard (ni-
`internal
`Fifty microliters of
`trazepam 5 mg:ml) methanol solution and 6 ml of
`ethyl acetate were added to 0.5 ml of the plasma,
`and the mixture was shaken for 10 min. After
`centrifugation (25°C, 3000 rpm, 10 min), 5ml of
`the ethyl acetate layer was taken and evaporated,
`and the residue was dissolved in 200 ml of 50%(v:
`v) acetonitrile aqueous solution. The concentra-
`tion of CZP was determined by HPLC.
`Apparatus and conditions of HPLC were the
`same as described in Section 3.3.2.
`
`4. Results and Discussion
`
`4.1. Identification of prepared samples
`
`Powder X-ray diffraction patterns of the com-
`mercial bulk (sample A), sample B% and sample C
`are shown in Fig. 2. Characteristic diffraction
`peaks were observed at 2u(cid:30)15.2, 15.8 and 17.0°
`for sample A, 2u(cid:30) 6.1, 9.4 and 19.9° for sample
`B% and 2u(cid:30)8.9, 18.9 and 19.4° for sample C. The
`diffraction patterns of sample A, sample B% and
`sample C agreed with those of form I, form III
`and dihydrate, respectively, given in the previous
`reports (Kaneniwa et al., 1984; Umeda et al.,
`1984).
`scanning calorimetry curves of
`Differential
`sample A, sample B% and sample C and TG curve
`of sample C are illustrated in Fig. 3. The DSC
`curve of sample A exhibited an endotherm at
`174°C followed by an exotherm at 176°C and a
`sharp endotherm at 190°C. The DSC curve of
`sample B% exhibited only one sharp endotherm at
`190°C. On the DSC curve of sample C, a broad
`endotherm at 50–75°C and a sharp endotherm at
`
`Fig. 2. Powder X-ray diffraction patterns of CZP samples. a:
`Sample A, b: Sample B%, c: Sample C.
`
`730(v:v), flow rate; 10 ml:min, column; Capcel-
`pack UG120 (4.6mm(cid:29)15cm, Shiseido) at 40°C,
`detection wavelength; 285 nm, injection volume;
`50 ml.
`
`3.2.3. Hygroscopicity studies
`Accurately weighed amounts of either sample A
`or sample B% were stored at 40°C and 98% relative
`humidity (RH). The 98% RH condition was pre-
`pared using saturated solution of ammonium
`phosphate in a desiccator. Weight changes of
`samples were monitored after 7, 14 and 28 days.
`
`3.3. Bioa6ailability tests
`
`3.3.1. Animal experiments
`Bioavailability studies were performed using a
`cross-over technique in four male beagle dogs
`which were fasted for at least 16 h before adminis-
`tration. The weights of dogs ranged from 9.2 to
`10.7 kg. Each sample was administered orally as a
`capsule with 40 ml of water. In addition, CZP
`solution using polyethyleneglycol 400 (PEG400)
`was administered orally. Blood was taken prior to
`administration and at 15, 30, 45 min and at 1, 2,
`3, 5, 8 and 12 h after dosing. Plasma was sepa-
`rated by centrifugation (4°C, 3000 rpm, 10 min)
`and stored in a freezer at (cid:28)20°C until analyzed.
`
`Merck Exhibit 2162, Page 4
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`Y. Kobayashi et al. :International Journal of Pharmaceutics 193 (2000) 137–146
`
`141
`
`Fig. 3. DSC and TG curves of CZP samples. a: DSC curve of Sample A, b: DSC curve of Sample B%, c: DSC curve of Sample C,
`d: TG curve of Sample C.
`
`where S is the surface area of the disk, V is the
`volume of test solution, k is the intrinsic dissolu-
`tion rate constant, and Cs is solubility. The disso-
`lution rate from the unit
`surface area,
`i.e.,
`intrinsic dissolution rate (IDR), is defined by Eq.
`(2).
`
`190°C were observed. On the TG curve of sample
`C, the weight loss corresponding to the DSC
`endotherm at 50–75°C was 13.1%, which was
`nearly equal to the stoichiometric value calculated
`for the dihydrate of CZP (13.2%). The results of
`thermal analysis were also consistent with those of
`form I, form III and dihydrate, respectively, given
`in the previous reports (Kaneniwa et al., 1984,
`Matsuda et al., 1984).
`These findings showed that sample A, B% and C
`were form I, form III and dihydrate, respectively.
`
`4.2. Physicochemical properties
`
`4.2.1. Dissolution studies by the static disk
`method
`Dissolution patterns of form I, form III and
`dihydrate in JP13 first fluid (pH 1.2) from 0 to 10
`min are shown in Fig. 4. Good linearities between
`time and concentration were found for each form.
`In the sink condition, the concentration of drug,
`C, at time t was expressed by Eq. (1) (Nogami et
`al., 1966).
`
`(1)
`
`Fig. 4. Dissolution patterns of CZP polymorphs and dibydrate
`determined by the static disk method in JP13 1st fluid at 37°C,
`(0–10 min),
`: form I, : form III, (cid:5): dihydrate.
`
`kCst
`
`S V
`
`C(cid:30)
`
`Merck Exhibit 2162, Page 5
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`142
`
`Y. Kobayashi et al. :International Journal of Pharmaceutics 193 (2000) 137–146
`
`since k depends on the diffusion coefficient and
`thickness of the diffusion layer.
`The intrinsic dissolution rate constant (k) was
`calculated by substitution of both the obtained
`solubility (described below) and IDR of dihy-
`drate. The value of k obtained was 0.134 cm:min.
`Solubilities of form I and form III, which were
`metastable forms in aqueous medium, could be
`calculated using k and IDR of form I and form
`III, respectively. The observed solubility of dihy-
`drate and the estimated solubilities of form I and
`form III are listed in Table 2. Comparing with the
`solubility of dihydrate, those of anhydrates were
`1.5–1.6 times higher. The solubility of form III
`was higher than that of form I, while the differ-
`ence in solubilities between the two anhydrates
`was relatively small. Kaneniwa et al. (1987) re-
`ported the difference in solubilities between anhy-
`drates and hydrate, but the order of solubility of
`form I and form III was inconsistent with that
`observed in the present study. Behme and Brooke
`(1991) reported that form I and form III were
`enantiotropic and that form I was stable below
`71°C. Therefore, the finding that the solubility of
`form I was less than that of form III appears to
`be reasonable.
`
`Table 2
`Intrinsic dissolution rates (IDR) and solubilities of CZP poly-
`morphs and dihydrate in JP13 1st fluid at 37°C
`
`IDR (mg:cm2:min)
`
`solubility (mg:ml)
`
`(2)
`
`Form I
`Form III
`Dihydrate
`
`61.8
`67.4
`41.8
`
`460.2a
`501.9a
`311.1b
`
`a Estimated value.
`b Determined by dispersion method.
`
`(cid:30)kCs
`
`V S
`
`C t
`
`IDR(cid:30)
`
`Intrinsic dissolution rates of form I, form III
`and dihydrate, which were calculated from the
`slope of each form in Fig. 4, surface area (S) and
`volume of test solution (V), were 61.8, 67.4 and
`41.8 mg:cm2:min (Table 2), respectively. Kaplan
`(1972) noted that compounds with IDR below 0.1
`mg:cm2:min usually exhibited problems with dis-
`solution rate-limited absorption. According to
`Kaplan’s classification, form I, form III and dihy-
`drate belonged to the category of dissolution rate-
`limited drugs.
`The intrinsic dissolution rate constants (k) of
`form I, form III and dihydrate must be equal,
`
`Fig. 5. Dissolution patterns of CZP polymorphs and dihydrate determined by the static disk method in JP13 1st fluid at 37°C,
`:
`form I, : form III, (cid:5): dihydrate.
`
`Merck Exhibit 2162, Page 6
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`Y. Kobayashi et al. :International Journal of Pharmaceutics 193 (2000) 137–146
`
`143
`
`the slopes of anhydrates after transformation to
`hydrate did not completely agree with that of
`dihydrate (Fig. 5). The inconsistency in dissolu-
`tion rates of dihydrate and anhydrates after trans-
`formation may have been due to changes in the
`surface area of anhydrate disks accompanied by
`dissolution of anhydrate and transformation to
`dihydrate.
`
`4.2.2. Dissolution studies by the dispersion
`method
`the individual
`The dissolution behaviors of
`form determined by the dispersion method in
`JP13 first fluid at 37°C are demonstrated in Fig. 6.
`Form m exhibited a rapid increase in CZP con-
`centration and the maximum value at the initial
`stage, then the concentration gradually decreased
`with precipitation of dihydrate crystals. On the
`other hand, form I maintained high constant con-
`centration over 4 h. These observations suggested
`that the transformation from form III to dihy-
`drate was faster than that of form I, which was
`compatible with the results obtained by using the
`static disk method.
`The maximum concentrations obtained for
`form I and form III, 459.3 and 499.5 mg:ml,
`respectively, were in good agreement with the
`solubilities estimated from the intrinsic dissolution
`rates of form I and form III (Table 2).
`
`4.2.3. Hygroscopicity studies
`The weight increase of form I and form m
`stored at 40°C and 98% RH for 28 days are
`exhibited in Fig. 7. Form III exhibited a gradual
`increase in weight due to water vapor adsorption,
`and after 2 weeks the water content of form III
`reached 13.7%, corresponding to the stoichiomet-
`ric value of the dihydrate (13.2%). On the other
`hand, the weight increase of form I after 28 days
`was only 1.9%. The crystalline forms of both form
`I and form III stored for 28 days were confirmed
`by powder X-ray diffractometry. For form III,
`the powder X-ray diffraction pattern was consis-
`tent with the pattern of the dihydrate, while for
`form I the pattern was the same as that before the
`test (Fig. 8). The results of hygroscopicity studies
`revealed that form III transformed to dihydrate
`faster than did form I.
`
`Fig. 6. Dissolution patterns of CZP polymorphs and dibydrate
`in JP13 1st fluid at 37°C,
`: form I, : form III, (cid:5):
`dihydrate.
`
`Fig. 7. Hygroscopicity study of polymorphs at 40°C and 98%
`RH,
`: form I, : form III.
`
`The dissolution curves of the individual form
`for prolonged time periods are shown in Fig. 5.
`Decreases in dissolution rates of form I and form
`III were observed after 50 and 20 min, respec-
`tively. Differential scanning calorimetry measure-
`ment of
`the surface of
`the disk after
`the
`dissolution test confirmed that both form I and
`form III were thoroughly transformed to dihy-
`drate. The changes in dissolution rates of anhy-
`drates were due to transformation to hydrate, and
`the rate of transformation from form III to dihy-
`drate was higher than that from form I. However,
`
`Merck Exhibit 2162, Page 7
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`144
`
`Y. Kobayashi et al. :International Journal of Pharmaceutics 193 (2000) 137–146
`
`A comparison of crystal data and molecular
`arrangement between form I and form III was
`performed by Roberts and Rowe (1996). Both
`form I and form III formed centrosymmetric
`dimers via intermolecular hydrogen bonding be-
`tween the carboxamide groups, however, the crys-
`tal systems differed; form I and form III were
`monoclinic and triclinic, respectively. Form I was
`more stacked than form III, in which significant
`hexagonal voids were included in the lattice due
`to three dimers forming a triangular structure.
`Consequently, the density of form III(1.24 g:cm3)
`was less than that of form I (1.34 g:cm3). The
`significant hexagonal void in form III might facil-
`itate the invasion of water molecules into the
`crystal lattice of form III and accelerate the trans-
`formation to dihydrate.
`
`4.3. Bioa6ailability tests
`
`Since the clinical dose of CZP in humans is
`increased up to la:day based on clinical efficacy,
`BA tests in dogs were carried out at a dose of
`
`either 40 mg:body or 200 mg:body. Plasma con-
`centration-time curves of form I, form III and
`dihydrate after oral administration at doses of 40
`mg:body and 200 mg:body in dogs are demon-
`strated in Fig. 9 and Fig. 10, respectively. Phar-
`macokinetic parameters are shown in Table 3.
`With the low dose, the plasma concentration-
`time curves of the three forms were nearly equal,
`and there was no significant differences in Cmax,
`Tmax, or AUC among the forms. The BA of each
`form was more than 80% of the AUC for PEG
`solution;
`this BA value was corresponding to
`relative BA. On the other hand, with the high
`dose, marked differences in plasma concentra-
`tion–time curves were found among the three
`forms. The plasma concentration level of form I
`was the highest among three forms, while those of
`form III and dihydrate were similar. The Cmax of
`form I was approximately twice those of form III
`and dihydrate. The AUC of the forms were in the
`order of form I\form III\dihydrate. The rela-
`tive BAs of form I and dihydrate were 68.7 and
`33.1%, respectively. However, this finding that
`
`Fig. 8. Powder X-ray diffraction patterns of CZP polymorphs stored at 40°C and 98% RH for 4 weeks. a: Intact form I, b: form
`I stored at 40°C and 98% RH for 4 weeks, c: intact form III, d: form III stored at 40°C and 98% RH for 4 weeks, e: dihydrate.
`
`Merck Exhibit 2162, Page 8
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`Y. Kobayashi et al. :International Journal of Pharmaceutics 193 (2000) 137–146
`
`145
`
`Fig. 9. Plasma concentration–time curves of CZP polymorphs
`and dihydrate after oral administration to dogs
`(n(cid:30)4;
`mean9S.E.). Dose: 40 mg:body, ": solution,
`: form I, :
`form III, (cid:5): dihydrate.
`
`Fig. 10. Plasma concentration–time curves of CZP poly-
`morphs and dihydrate after oral administration to dogs (n(cid:30)4;
`mean9S.E.). Dose: 200 mg:body, ": solution,
`: form I, :
`form III, (cid:5): dihydrate.
`
`Table 3
`Pharmacokinetic parameters of CZP polymorphs and dihydrate after oral administration in dogs (n(cid:30)4; mean9S.E.)*
`
`Dose:40 mg:body
`Solution
`Form I
`Form III
`Dihydrate
`
`Dose: 200 mg:body
`Solution
`Form I
`Form III
`Dihydrate
`
`Cmax (mg:ml)
`
`1.0690.17a
`0.7190.16a
`0.7090.05a
`0.8090.06a
`
`5.5990.45a
`4.2990.41b
`2.3690.65c
`1.9090.38c
`
`Tmax (h)
`
`0.490.1a
`0.890.2a
`0.790.1a
`0.990.1a
`
`0.690.1a
`1.190.4a
`1.390.4a
`0.890.1a
`
`AUC (mg.h :ml)
`
`Relative BA (%)
`
`1.7690.20a
`1.5390.07a
`1.4690.18a
`1.5990.18a
`
`13.2591.20a
`9.1091.00b
`6.3392.39b, c
`4.3991.30c
`
`100.0a
`86.9a
`82.9a
`90.3a
`
`100.0a
`68.7 b
`47.8 b’C
`33.1c
`
`* Different superscripts indicate significant differences among the means (PB0.05) and same superscripts indicate no significant
`difference.
`
`form I had higher BA than form III, was inconsis-
`tent with the thermodynamic stability.
`The similarity in BA of the three crystal forms
`when administered at a dose of 40 mg:body could
`be explained by the dissolution behavior of each
`form; most of the administered samples dissolved
`in the gastrointestinal tract. This consideration
`was also supported by the high relative BAs for
`
`each form as compared with that for PEG solu-
`tion. On the other hand, with the high dose (200
`mg:body), the dissolution property of each form
`might have affected the plasma concentration of
`CZP, since most of
`the administered powder
`could not dissolve in the GI tract due to low
`solubility and excess dose. The finding of lowest
`solubility of dihydrate among the three forms
`
`Merck Exhibit 2162, Page 9
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`146
`
`Y. Kobayashi et al. :International Journal of Pharmaceutics 193 (2000) 137–146
`
`could account for the finding of lowest AUC of
`dihydrate. The differences in BA and thermody-
`namic stability between form I and form m should
`be attributed to the more rapid transformation
`from form III to dibydrate than that from form I.
`This consideration for the relationship between
`BAs of form I and form III was supported by the
`similar plasma concentration–time profiles of
`form m and dihydrate as administered orally at
`high dose. The S.E.s of Cmax and AUC for form
`III were larger than those for the other forms.
`The scatter of form III parameters was due to
`variability in effects on transformation from form
`III to dihydrate of motility, pH and volume of the
`individual GI tract.
`In this study, results of evaluation of BA for
`form I, form III and dihydrate CZP in dogs
`differed between doses administered (40 mg:body
`and 200 mg:body). Assessment of BA for water-
`insoluble compounds with polymorphism or pseu-
`dopolymorphism should be performed with
`various doses.
`
`References
`
`Behme, R.J., Brooke, D., 1991. Heat of fusion measurement of
`a low melting polymorph of carbamazopine that undergoes
`multiple phase
`changes during differential
`scanning
`calorimetry analysis. J. Pharm. Sci. 80, 986–990.
`Ito, S., Nishimura, M., Kobayashi, Y., Itai, S., Yamamoto,
`K., 1997. Characterization of polymorphs and hydrates of
`GK-128, a serotonin3 receptor antagonist. Int. J. Pharm.
`151, 133–143.
`
`Kahela, P., Aaltonen, R., Lewing, E., Anttila, M., Kristof-
`fersson, E., 1983. Pharmacokinetics and dissolution of two
`crystalline forms of carbamazepine. Int. J. Pharm. 14,
`103–112.
`Kaneniwa, N., Ichikawa, J., Yamaguchi, T., Hayashi, K.,
`Watari, N., Sumi, M., 1987. Dissolution behabior of car-
`bamazepine polymorphs. Yakugaku Zasshi 107, 808–813.
`Kaneniwa, N., Yamaguchi, T., Watari, N., Otsuka, M., 1984.
`Hygroscopicity of carbamazepine crystalline powders.
`Yakugaku Zasshi 104, 184–190.
`Kaplan, S.A., 1972. Biopharmaceutical considerations in drug
`formation design and evaluation. Drug Metab. Rev. 1,
`15–34.
`Krahn, F.U., Mielck, J.B., 1987. Relations between several
`polymorphic forms and the dihydrate of carbamazepine.
`Pharm. Acta Helv. 62, 247–254.
`McMahon, L.E., Timmins, P., Williams, A.C., York, P., 1996.
`Characterization of dihydrates prepared from carba-
`mazepine polymorphs. J. Pharm. Sci. 85, 1064–1069.
`Meyer, M.C., Straughn, A.B., Jarivi, E.J., Woods, G.C., Pel-
`sor, F.R., Shah, V.P., 1992. The bioinequivalence of carba-
`mazepine tablets with a history of clinical failures. Pharm.
`Res. 9, 1612–1616.
`Nogami, H., Nagai, T., Suzuki, A., 1966. Studies on powdered
`preparations. XVII. Dissolution rate of sulfonamides by
`rotating disk method. Chem. Pharm. Bull. 14, 329–338.
`Rajendra, K.K., David, J.W.G., 1995. Pharmaceutical hy-
`drates. Thermochimica Acta 248, 61–79.
`Roberts, R.J., Rowe, R.C., 1996. Influence of polymorphism
`on the Young’s modulus and yield stress of carbama-
`zepine, sulfathiazole and sulfanilamide. Int. J. Pharm. 129,
`79–94.
`Umeda, T., Ohnishi, N., Yokoyama, T., Kuroda, K., Kuroda,
`T., Matsuda, Y., 1984. Kinetics of the thermal transition of
`carbamazepine polymorphic forms in the solid state. Yaku-
`gaku Zasshi 104, 786–792.
`York, P., 1983. Solid-state properties of powders in the formu-
`lation and processing of solid dosage forms. Int. J. Pharm.
`14, 1–28.
`
`Merck Exhibit 2162, Page 10
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`